AI Usage by Students in Business Information Systems Courses: An Early Survey

Abstract

California State University is the largest state university system in the United States. Recently, it announced a landmark initiative to become the nation's first and largest AI-empowered university system. The first of its kind public-private initiative involved Adobe, Alphabet, AWS, IBM, Instructure, Intel, LinkedIn, Microsoft, NVIDIA, and OpenAI. AI tools and training are now available to nearly 460,000 students and 63,000 faculty members. Although the initiative started in February 2025, the adoption of AI progressed rapidly in many of the California State University campuses. This paper describes the use of AI tools in business core courses offered by the Information Systems Department at the College of Business and Economics at California State University, Los Angeles. It also discusses students' adoption of AI tools based on informal observations and a survey conducted at the end of the Spring 2025 semester. The survey questions focused on the following: chatbots used frequently, activities for which students use chatbots, different AI tools used by students, challenges faced when using a chatbot, rating of chatbots, concerns about AI tools, use of chatbots for technical applications, use of AI tools in other classes, and effectiveness of AI tools. The survey and informal observation indicated that the students initially used the chatbots for advanced AI-aided Internet search, brainstorming, report writing, and project assignments. As they became more familiar with the AI tools, it became evident that they could generate study guides and multiple-choice practice questions to prepare for quizzes and examinations with AI. The students started summarizing PowerPoint slides and lecture notes to help them learn. AI tools thus enabled them to devise self-study pathways to enhance their learning experience. Although the students readily adopted the AI tools for many applications, they were slow to use the tools for more computationally intensive tasks, such as manipulating data in spreadsheets. The slow adoption may be attributed to chatbots such as Copilot being available in the student version of chatbots as a stand-alone AI tool. Only the pro version of Copilot offered the option to integrate Copilot as an add-on within Excel. Another technical application of ChatGPT where students were slow to adopt was computer modeling and code generation. The adoption of chatbots for computer modelling and code generation was slow because both required a good understanding of the models and excellent prompt engineering skills. Overall, the students at the College of Business and Economics embraced the AI tools as quickly as possible. At the same time, they felt that AI tools were ineffective in helping them with critical thinking. AI tools are generative and not creative. Critical thinking is necessary to augment generative AI tools. Therefore, instructors designing curriculum in the future must give primacy to critical thinking and prompt engineering. Finally, contrary to popular sentiments, as the initiator and curator of learning, the instructor will continue to be at the center of instructional pedagogy, including curriculum design and instruction delivery.